This thread may be a little bit of old news by now, but thanks to those 
who responded here and on the Duke Digital Collections blog.  I'm sort 
of getting back to it now after a conference trip, and a week spent on a 
major update to our digital collections site.

Based on the responses, I conclude (A being an anecdotal response, N the 
number of responses, and D data, thus AN = D) that the environment has 
evolved since my inquiry in January 2007.  People are still using 
spreadsheets and other desktop productivity applications to capture 
non-MARC metadata.  But more places have taken on development of a 
platform that scales to their own institutional needs.  There are also 
some very early attempts to release these projects for reuse and 
collaboration with peer institutions.

There seems to be a fair amount of traction in this area right now.  
I'll keep this community informed of our progress and hope others will 
do the same.

Will
--
Will Sexton
Metadata Analyst / Programmer
Duke University Libraries